A Detection Method for Bearing Faults of Marine Motors Based on Data Mining Algorithm

This paper proposes an improved CLIQUE algorithm for detection of marine motor´s bearing faults. The major theoretical principles of the algorithm based on spectral analysis are described. The presented approach works simply and is thus suited for condition monitoring of machine system under varying operating and loading conditions. The performance of this technique is investigated through study of realistic current signals. Experimental results show that the proposed method has better performance and validity in realizing bearing faults of marine motors. The limitation to extract the fault characteristic frequency resulting from the fluctuation of the characteristic frequency and the variation of the load is overcome.